Jeesu Choe, Boyoung Oh, and Eunok Choe
Abstract: The effect of soybean lecithin addition on the iron-catalyzed or chlorophyll-photosensitized oxidation of emulsions consisting of purified canola oil and water (1:1, w/w) was studied based on headspace oxygen consumption using gas chromatography and hydroperoxide production using the ferric thiocyanate method. Addition levels of iron sulfate, chlorophyll, and soybean lecithin were 5, 4, and 350 mg/kg, respectively. Phospholipids (PLs) during oxidation of the emulsions were monitored by high performance liquid chromatography. Addition of soybean lecithin to the emulsions significantly reduced and decelerated iron-catalyzed oil oxidation by lowering headspace oxygen consumption and hydroperoxide production. However, soybean lecithin had no significant antioxidant effect on chlorophyll-photosensitized oxidation of the emulsions. PLs in soybean lecithin added to the emulsions were degraded during both oxidation processes, although there was little change in PL composition. Among PLs in soybean lecithin, phosphatidylethanolamine and phosphatidylinositol were degraded the fastest in the iron-catalyzed and the chlorophyll-photosensitized oxidation, respectively. The results suggest that addition of soybean lecithin as an emulsifier can also improve the oxidative stability of oil in an emulsion. Keywords: chlorophyll-photosensitized oxidation, degradation of phospholipids, emulsion, iron-catalyzed oxidation, soybean lecithin
The results of this study can be applied to extend the utilization of soybean lecithin for improvement of the oxidative stability of emulsions as well as an emulsifier.
Practical Application:
Introduction The quality of an emulsion is largely determined by the oxidation of oil in the system (Lee and Choe 2011), and the oil oxidation is accelerated by transition metals such as iron as well as pigments such as chlorophylls under light (An and others 2011). An emulsion is stabilized by an emulsifier such as phospholipids (PLs), which have both a polar head and a nonpolar tail in their structures. Until now, there remains a controversy regarding the effect of PL on oil oxidation. Phosphatidylcholine (PC) reduces the autoxidation of bulk oil (Sugino and others 1997; Zambiazi and Przybylski 1998; Koo and Kim 2005) by chelating or binding to prooxidant metal ions (Jacobson and Papahadjopoulos 1975; Viani and others 1990). On the other hand, PL worked as a prooxidant in both autoxidation (Husain and others 1986; Yoon and Min 1987) and photooxidation of bulk oil (Lee and Choe 2009) by increasing oxygen diffusion from the headspace to the oil or protecting chlorophylls, photosensitizers for more reactive singlet oxygen production under light, from photodegradation. 1,2-Dioleoyl-sn-glycerol-3-phosphocholine forming reverse micelles in stripped soybean oil acted as a prooxidant, whereas 1,2-dibutyl-sn-glycerol-3-phosphocholine was not able to form association colloids and did not significantly affect lipid oxidation (Chen 2012). It is generally agreed that the effect of PLs on the oil oxidation is dependent on the type and the concentration of
PLs, a reaction system matrix whether bulk oil or an emulsion, and the presence of iron or other antioxidants (Judde and others 2003; Choe and Min 2006). Most research on the effect of PLs on oil oxidation has examined autoxidation using pure PLs such as pure PC or phosphatidylethanolamine (PE; Jacobson and Papahadjopoulos 1975; Yoon and Min 1987; Sugino and others 1997; Lee and Choe 2009). Lecithin, a common emulsifier used in the food industry, is a mixture of PLs including PC, PE, and phosphatidylinositol (PI), and is naturally found in egg yolk and soybeans. Due to the high cost of lecithin from egg yolk, soybean lecithin is more widely used in the food industry. Since the PL and fatty acid composition of lecithin varies depending on the sources, lecithin extracted from different food materials may have different effects on oil oxidation in an emulsion. Another concern in terms of oil oxidation in emulsion foods with added PLs may be light exposure during transportation and storage, and a photosensitized-oxidation may occur if photosensitizing pigments are present. This research evaluated an oil oxidation and PL degradation in emulsions with added soybean lecithin under iron or light with chlorophyll in order to elucidate the effect and action mechanism of soybean lecithin on iron-catalyzed or chlorophyll-photosensitized oxidation of canola oil emulsions.
Materials and Methods MS 20140955 Submitted 6/3/2014, Accepted 9/5/2014. Authors are with Dept. of Food and Nutrition, Inha Univ., Incheon, South Korea. Direct inquiries to author Choe (E-mail: [email protected]).
R C 2014 Institute of Food Technologists
doi: 10.1111/1750-3841.12683 Further reproduction without permission is prohibited
Materials and reagents Refined, bleached, and deodorized (RBD) canola oil was obtained from CJ Co. (Seoul, Korea), and soybean lecithin was Vol. 79, Nr. 11, 2014 r Journal of Food Science C2203
C: Food Chemistry
Effect of Soybean Lecithin on Iron-Catalyzed or Chlorophyll-Photosensitized Oxidation of Canola Oil Emulsion
Lecithin on emulsion oxidation . . .
C: Food Chemistry
purchased from Sigma-Aldrich Co. (St. Louis, Mo., U.S.A.). n-Hexane, methanol, acetonitrile, and chloroform were high performance liquid chromatography (HPLC) grade and purchased from Mallinckrodt Baker Co. (Phillipsburg, N.J., U.S.A.). Chlorophyll b, 14% BF3 in methanol, silicic acid, alumina, xanthan gum, standard PC, PE, PI, phosphatidylserine (PS), heptadecanoic acid, methyl esters of standard fatty acids (C4–C24 saturated fatty acids [SFAs] as well as oleic, linoleic, and α-linolenic acids), and ammonium thiocyanate were obtained from Sigma-Aldrich Co. Isooctane, celite, and ferrous sulfate were purchased from Junsei Co. (Tokyo, Japan). All other chemicals were analytical grade.
Preparation of emulsions Emulsion samples were prepared with tocopherol-stripped canola oil (TSCO), distilled water, and xanthan gum according to the method of Lee and Choe (2011). Canola oil was chosen due to its relatively high content of linolenic acid (11.48 ± 0.05%), one of the omega-3 fatty acids. TSCO was obtained by passing the RBD canola oil through a glass column (30 cm × 2.5 cm) packed with activated silicic acid, alumina, and celite from the bottom (Lee and Choe 2009). TSCO (49.8 g), water (49.8 g), and xanthan gum (0.35 g) were mixed at room temperature, followed by addition of FeSO4 or chlorophyll b, and soybean lecithin. Concentration of FeSO4, chlorophyll b, and soybean lecithin were 5, 4, and 350 mg/kg, respectively, which were within the range to show significant effect on the oil oxidation in emulsions (Osborn and Akoh 2003; Lee and Choe 2011). The mixture was then homogenized for 6 min using an Ultra-Turrax T25 homogenizer equipped with an S25N-25F dispersing tool (IKA Instruments, Staufen, Germany). Control samples did not contain soybean lecithin, but the emulsion system was maintained for 10 d similar to other samples, possibly due to xanthan gum. The emulsion (3 g) was transferred into 20-mL serum bottles, which were then tightly capped with rubber stoppers and aluminum caps. Samples with added iron were covered with aluminum foil to exclude light. All samples in bottles were oxidized in an LBI-250 incubator (Daihan Labtech Co., Seoul, Korea) under 1700 lux light and maintained at 25 °C for 10 d until taken out for analyses. All samples were prepared in duplicate. Evaluation of oil oxidation in emulsions Oil oxidation in the emulsions was determined based on the headspace oxygen consumption as a loss of reactant (oxygen) in the oil oxidation, and the hydroperoxide contents were also determined as a primary product of oil oxidation. Monitoring of oxygen consumption is a very useful tool for the analysis of oil oxidation in an emulsion since it does not require extracting the oil phase before analysis and focuses on the loss of substrate in the oxidation reactions (Kristinov´a and others 2009). It also enables continuous monitoring of the oxygen concentration in the system and easy changes of environmental conditions such as light as well as addition of prooxidants and antioxidants to the system during the oxidation process. Thus, the oxygen consumption has been employed to provide direct information on the oil oxidation in emulsions (Mozuraityte and others 2006; Mozuraityte and others 2008; Carvajal and others 2009; Kristinov´a and others 2009; An and Choe 2011; Kristinova and others 2014). Headspace oxygen content was determined by gas chromatography (GC; Lee and Choe 2011) using a YL 6100 gas chromatograph (Younglin Instrument Co., Ltd., Anyang, Korea) equipped with an autosampler and a thermal conductivity detector. The injection volume was 0.5 mL. A column was the stainless steel column packed with C2204 Journal of Food Science r Vol. 79, Nr. 11, 2014
80/100 mesh molecular sieve 13× (1.83 m × 0.32 cm; Alltech, Deerfield, Ill., U.S.A.). Helium (99.995%) was a carrier gas at 20 mL/min. Temperatures of the oven, injector, and detector were 35, 100, and 140 °C, respectively. Areas of the oxygen peak in the chromatograms were converted to μmol of oxygen in 1 mL of headspace gas with a conversion factor of 9.35 (Lee and Choe 2011). Hydroperoxide content of the emulsion was determined by the ferric thiocyanate method (An and Choe 2011). After mixing the emulsion with isooctane and 2-propanol mixture (3:1, v/v), the organic layer was separated by centrifugation at 1000 × g for 2 min, followed by addition of methanol and chloroform mixture (2:1, v/v), 3.94 mol/L ammonium thiocyanate, 0.132 mol/L BaCl2 , and 0.144 mol/L FeSO4 solution in the respective order. The absorbance at 510 nm using a HP8453 UV-visible spectrophotometer (Hewlett Packard, Wilmington, Del., U.S.A.) was read after 20 min. Hydroperoxide content of oil was expressed with cumene hydroperoxide (CuOOH), and the range of linearity was between 0 and 10 mmol/kg (r2 = 0.9994).
Compositional analyses of soybean lecithin PL composition of soybean lecithin was analyzed by HPLC (Choe and Choe 2013). Soybean lecithin was dissolved in chloroform and then injected into a YL 9100 HPLC (Younglin Instrument Co., Ltd.) equipped with a UV detector at 205 nm. The injection volume was 20 μL. A cosmosil-packed column (5SL-II; 4.6 mm × 250 mm, 5 μm size; Nacalai Tesque, Kyoto, Japan) was used with an eluting solvent of acetonitrile, methanol, and phosphoric acid (130:5:1.5, v/v/v) at 1.0 mL/min in an isocratic manner. Each PL in soybean lecithin was identified by comparing the retention time with those of standard PC, PE, PI, and PS, and quantified with their respective calibration curves with a linearity range between 0 and 400 mg/L (r2 > 0.9950). Fatty acid composition of soybean lecithin was analyzed by GC according to the method of Lee and Choe (2011) after the lecithin was saponified with methanolic NaOH (2 mol/L) and esterified with 14% BF3 in methanol. The instrument was a YL 6100 gas chromatograph (Younglin Instrument Co., Ltd.) equipped with an autosampler, an HP-Innowax column (30 m × 0.53 mm, 1.0-μm thick; Agilent, B¨oblingen, Germany), and a flame ionization detector. The injection volume was 1 μL. The temperatures of the oven, injector, and detector were 200, 270, and 280 °C, respectively. The helium flow rate was 10 mL/min, and the split ratio was 10:1. Heptadecanoic acid was used as an internal standard, and each fatty acid in the chromatogram was identified by comparing retention times with those of standard fatty acid methyl esters. Concentration of each fatty acid was represented as relative % by dividing each peak area by total peak areas. Determination of PL contents in emulsions during oxidation Content of each PL in the emulsion during 10-d oxidation was determined by HPLC after phase separation. The emulsion was frozen, thawed, and centrifuged, and both interfacial and water phases were dissolved in n-hexane. The upper layer obtained by centrifugation (2000 × g, 10 min) was applied to a solid phase microextraction Sep-pak silica plus long cartridge (55–105 μm; Waters Corp. Milford, Mass., U.S.A.), and eluted with chloroform and methanol in a respective order (Lee and Choe 2011). PLs eluted with methanol were analyzed by HPLC using a YL 9100 HPLC (Younglin Instrument Co., Ltd.) under the same analytical conditions described previously.
Lecithin on emulsion oxidation . . .
PL content (mg/100 mg lecithin)
Fatty acid compositionb
Phosphatidylcholine
Phosphatidylethanolamine
Phosphatidylinositol
Phosphatidylserine
Total
31.9 ± 0.2 (40.7)a
26.4 ± 0.1 (33.6)
18.0 ± 0.2 (23.0)
2.1 ± 0.0 (2.7)
78.4 ± 0.1 (100)
Palmitic acid
Stearic acid
Oleic acid
Linoleic acid
Linolenic acid
USFA/SFAc
21.2 ± 0.2
4.2 ± 0.1
16.2 ± 0.1
53.4 ± 0.1
5.1 ± 0.1
3.0 ± 0.0
a
Relativity in % based on total PL content. b Relativity in % based on total fatty acids. c Content ratio of unsaturated fatty acids to saturated fatty acids.
Data analysis All samples were prepared in duplicate, and each sample was measured twice, resulting in 4 measurements for each treatment. Data were analyzed using one-way analysis of variance and Duncan’s multiple range test using SAS (version 9.2, SAS Inst. Inc., Cary, N.C., U.S.A.) at α = 0.05.
Results and Discussion
12
7 a b
6 c c
5
a
Hydroperoxide contents mmol CuOOH/ kg
Headspace oxygen consumption μmol O2 / mL
Composition of soybean lecithin PL content of soybean lecithin was 78.4%, and PL consisted of mainly PC (40.7%), PE (33.6%), and PI (23.0%) with a small amount of PS (2.7%), as shown in Table 1. These values were within the ranges reported by Garti (2002). Linoleic acid (53.4%) was the most abundant fatty acid in soybean lecithin along with palmitic (21.2%) and oleic (16.2%) acids. Linolenic (5.1%) and stearic (4.2%) acids were present at low concentration, resulting in the content ratio of unsaturated fatty acids (USFA) to SFAs of 3.0. High unsaturation of soybean lecithin was reported elsewhere (Szuhaj 1989), and high unsaturation of PL was suggested for its prooxidant effect in the oxidation of chicken fat (Pikul and Kummerow 1990).
Effect of soybean lecithin addition on oil oxidation in canola oil emulsions Headspace oxygen consumption and hydroperoxide contents of the canola oil emulsions increased during both iron-catalyzed autoxidation and chlorophyll-photosensitized oxidation at 25 °C for 10 d, as shown in Figure 1. The control emulsion without added soybean lecithin showed oxygen consumption levels of 1.14, 3.62, and 5.41 μmol O2 /mL as well as hydroperoxide contents of 2.55, 4.66, and 5.71 mmol CuOOH/kg after 2, 6, and 10 d of iron-catalyzed oxidation, respectively. Oxygen consumption and hydroperoxide contents of the emulsion with added soybean lecithin after 10 d of iron-catalyzed oxidation were significantly reduced to 3.32 μmol O2 /mL and 3.16 mmol CuOOH/kg, respectively, which were significantly lower than those of the control emulsion (P < 0.05). These results indicate that soybean lecithin significantly reduced oil oxidation catalyzed by iron in the canola oil emulsions. Added soybean lecithin has been reported to show antioxidant activity during heating of vegetable oil at 110 °C (Judde and others 2003), whereas Fomuso and others (2002) reported increased autoxidation of fish oil-based structured lipid emulsion by addition of lecithin.
c d
4 ef f
3
e ef g
g 2
h i
1 j 0
h i
j
kj 0
2
4 6 Oxidation time days
8
10
a
10
b 8
bc
c d
6
e
e
f 4 gh gh h
2
i
i
0 0
2
g
g gh h
h i 4 6 Oxidation time days
8
10
Figure 1–Effect of soybean lecithin (350 mg/kg) on the headspace oxygen consumption and hydroperoxide contents of canola oil emulsions during iron-catalyzed autoxidation (solid line) or chlorophyll-photosensitized oxidation (dotted line) at 25 °C (x, no soybean lecithin addition (control); , with soybean lecithin addition). Different letters in lines represent significant difference among values regardless of oxidation time and treatment in each oxidation parameter by Duncan’s multiple range test at α = 0.05. Vol. 79, Nr. 11, 2014 r Journal of Food Science C2205
C: Food Chemistry
Table 1–Composition of soybean lecithin.
Lecithin on emulsion oxidation . . . Table 2–Effect of soybean lecithin addition on regression analysis between oxidation time and headspace oxygen consumption or hydroperoxide contents of canola oil emulsions during iron-catalyzed autoxidation or chlorophyll-photosensitized oxidation at 25 °C for 10 d. Oxidation
Regression parametersa
Soybean lecithin addition
Headspace oxygen consumption (μmol O2 /mL)
C: Food Chemistry
Iron-catalyzed Chlorophyll-photosensitized
No (control) Yes No (control) Yes
Hydroperoxide contents (mmol CuOOH/kg)
a
b
r2
a
b
r2
0.556 ab Ac 0.340 bb 0.710 ab Ac 0.685 ab
0.098 −0.198 0.190 0.498
0.994 0.989 0.947 0.955
0.582 ab Ac 0.319 bb 0.976 ab Ac 0.908 ab
0.535 −0.335 0.245 0.447
0.940 0.918 0.864 0.901
Headspace oxygen consumption (μmol O2 /mL) or hydroperoxide contents (mmol CuOOH/kg) = a × oxidation time (d) + b, r2 = determination coefficient Different lowercase letters represent a significant difference in the slope between control and sample with added soybean lecithin under each oxidation by dummy variable regression analysis (P < 0.05). c Same uppercase letters represent no significant difference in the slope between control samples under iron-catalyzed autoxidation and chlorophyll-photosensitized oxidation by dummy variable regression analysis (P < 0.05). a
b
Oxygen consumption and hydroperoxide production of the control emulsion under chlorophyll-photosensitized oxidation were significantly higher than those under iron-catalyzed autoxidation (P < 0.05); headspace oxygen consumption of the control emulsion without soybean lecithin addition after 6 and 10 d of oxidation was 5.46 and 6.53 μmol O2 /mL, while hydroperoxide content was 6.63 and 8.16 mmol CuOOH/kg, respectively. This result clearly confirms higher oil oxidation in the emulsions under light in the presence of chlorophyll than in the dark in the iron presence. Chlorophyll acts as a photosensitizer for the production of singlet oxygen, whose reactivity with lipids is approximately 1400 times higher than that of triplet oxygen (Choe and Min 2006). Unlike iron-catalyzed oxidation, addition of soybean lecithin did not significantly reduce oxygen consumption or hydroperoxide production in the emulsions under chlorophyll-photosensitized oxidation. Oxygen consumption and hydroperoxide content of the emulsion with added soybean lecithin after 2, 6, and 10 d of oxidation were 1.92 μmol O2 /mL and 2.47 mmol CuOOH/kg, 5.19 μmol O2 /mL and 7.57 mmol CuOOH/kg, and 6.52 μmol O2 /mL and 8.22 mmol CuOOH/kg, respectively, which were not significantly different from those of the control emulsion (P > 0.05). These results indicate that soybean lecithin did not affect chlorophyll-photosensitized oxidation of the canola oil emulsions. Until now, the effect of PLs on chlorophyll-photosensitized oxidation in emulsions has not been reported. Headspace oxygen consumption and hydroperoxide production in the canola oil emulsions were highly correlated with the oxidation time (r2 > 0.86), as shown in Table 2. Rates of oil oxidation were estimated based on the slope of the regression lines between time and oxygen consumption or hydroperoxide content, as previously described (Kim and Choe 2012). The rate of oil oxidation in the control emulsion was significantly higher (P < 0.05) under chlorophyll-photosensitized oxidation (0.710 μmol O2 /mL/d and 0.976 mmol CuOOH/kg/d) than under iron-catalyzed oxidation (0.556 μmol O2 /mL/d and 0.582 mmol CuOOH/kg/d). Further, the rate of oil oxidation was significantly lower (P < 0.05) in the emulsion with added soybean lecithin (0.340 μmol O2 /mL/d and 0.319 mmol CuOOH/kg/d) than in the control emulsion only under iron-catalyzed autoxidation. In the presence of chlorophyll under light, the rate of oil oxidation in the emulsion with added soybean lecithin was 0.685 μmol O2 /mL/d based on oxygen consumption and 0.908 mmol CuOOH/kg/d based on the hydroperoxide production, and these were not significantly different from the rate of oil oxidation in the
C2206 Journal of Food Science r Vol. 79, Nr. 11, 2014
control emulsion (P > 0.05). PLs sequester metals and/or donate hydrogen atoms from their nitrogen-containing functional groups such as choline or ethanolamine to radicals (Saito and Ishihara 1997), resulting in decreased oil oxidation (Choe and Min 2006). On the other hand, PLs can elevate autoxidation and chlorophyllphotosensitized oxidation of purified oil which does not have any metals (Yoon and Min 1987; Lee and Choe 2009). Since PLs are located mainly at interfaces with their nonpolar fatty acid tails facing the oil phase and a polar head facing the water phase (Xu and others 2011), they are able to decrease the surface tension of oil and increase oxygen diffusion from the headspace, resulting in acceleration of oil oxidation (Choe and Min 2006). Therefore, the reduction of iron-catalyzed autoxidation of oil by added soybean lecithin could have been largely attributed to iron chelation by PLs. Iron increases oil oxidation by accelerating decomposition of hydroperoxides into radicals (Choe and Min 2006). In addition, donation of hydrogen to radicals from functional groups such as choline or ethanolamine could have contributed to the reduction of oil oxidation. Although PLs can also donate hydrogens to radicals under chlorophyll-photosensitized oxidation, increased oxygen diffusion by PLs could have resulted in a higher singlet oxygen production, which was more predominant than radical scavenging by hydrogen donation in chlorophyll-photosensitized oxidation. Thus, addition of soybean lecithin had no significant effect on the chlorophyll-photosensitized oxidation of the canola oil emulsions.
PL contents in emulsions with added soybean lecithin during oxidation PL content of the emulsion with added soybean lecithin was measured to be 252.76 mg/kg before oxidation, which was lower than the expected level (274.4 mg/kg) since soybean lecithin with 78.40% of purity was added at 350 mg/kg. This suggests that some of PL in soybean lecithin were lost during PL analysis, with 92.1% of recovery. Content of PC (100.88 mg/kg), PE (90.26 mg/kg), and PI (61.62 mg/kg) corresponded to 39.9, 35.7, and 24.4%, respectively, of total PL content. There was no PS detected in the emulsions, possibly due to a very low amount compared to other PLs. During iron-catalyzed or chlorophyllphotosensitized oxidation at 25°C, PL contents in the canola oil emulsions significantly decreased (Figure 2), indicating PL degradation. However, PL composition changed little, with a slight decrease in the relative content of PE and an increase in PI. After 10 d of oxidation catalyzed by iron, contents of PC, PE, and PI
Lecithin on emulsion oxidation . . . Table 3–Degradation of phospholipids in canola oil emulsions with added soybean lecithin during iron-catalyzed autoxidation or chlorophyll-photosensitized oxidation at 25 o C for 10 d.
Iron-catalyzed
Regression parametersa
Phosphatidylcholine Phosphatidylethanolamine Phosphatidylinositol Total Phosphatidylcholine Phosphatidylethanolamine Phosphatidylinositol Total
Chlorophyll-photosensitized
k
t1/2
r2
0.065bb 0.094a 0.044bc 0.068Ac 0.053bb 0.046bc 0.078a 0.056Ac
10.664 7.374 15.753 10.193 13.078 15.068 8.887 12.378
0.9704 0.8940 0.8257 0.9656 0.9462 0.9061 0.9953 0.9723
Estimated by regression assuming 1st-order kinetics, ln([A]/[A]0 = −k × time (d), where [A] and [A]0 are contents of phospholipids (mg/kg) at times t and 0, respectively. r2 , determination coefficient; t1/2 , half-life. b Different letters (a, b, c) represent significant difference in degradation rate among phospholipid classes under each oxidation by dummy variable regression analysis (P < 0.05). c Same letters (A, A) represent no significant difference in degradation rates of total phospholipids between iron-catalyzed autoxidation and chlorophyll-photosensitized oxidation by dummy variable regression analysis (P < 0.05). a
were 51.51, 38.04, and 37.33 mg/kg, respectively, corresponding to 40.6%, 30.0%, and 29.4% of the initial content, giving the total PL content of 126.88 mg/kg (50.2% of the initial content). Degradation of PLs is related to their antioxidant activity by hydrogen donation to lipids (Yoshida and others 1995), and PC and PE were reported to be degraded into carbonyl groups and pyrrole compounds during autoxidation of soybean oil (Hidalgo and others 2005). PLs in the canola oil emulsions with added soybean lecithin were also degraded during chlorophyll-photosensitized oxidation at 25 °C (Figure 2). After an oxidation catalyzed by chlorophyll under light, contents of PC, PE, and PI significantly decreased to 41.3%, 35.5%, and 23.2% (after 2 d), 39.3%, 39.3%, and 21.4% (after 6 d), and 41.5%, 38.2%, and 20.3% (after 10 d) of the initial content, respectively. Total PL contents decreased to 234.07, 180.04, and 142.86 mg/kg after 2, 6, and 10 d of oxidation,
respectively, however, PL composition of the emulsions changed little, similar to the case of iron-catalyzed oxidation. Degradation of PLs in the canola oil emulsion with added soybean lecithin during iron-catalyzed or chlorophyll-photosensitized oxidation followed a 1st-order kinetics (r2 > 0.82). Rate constants (k value) for the 1st-order degradation of PLs were estimated by regression of the natural log (ln) of the PL content ratio ([A]t /[A]0 , where [A]0 and [A]t are PL contents at oxidation times 0 and t, respectively) against time, as shown in Table 3. A rate constant for the degradation of total soybean PLs during iron-catalyzed oxidation of the emulsion was 0.068/d, and the rate constants for PC, PE, and PI degradation were 0.065, 0.094, and 0.044/d, respectively. Among soybean PLs, degradation of PE was significantly faster (P < 0.05) than that of PC or PI during iron-catalyzed oxidation. Lower stability of PE compared to PC was reported during autoxidation of wheat bran oil (Choi and Choe 2009)
B Chlorophyllphotosensitized oxidation
A Iron-catalyzed autoxidation
350
350 300
a
250
b
200
c
cd f
150
g
a 100 50
Phospholipid contents mg/ kg
Phospholipid contents mg/ kg
300
a ab
b a ab
c
c
c
c cd cd
d cd
f
d g ef
a
250
b
200
c
d
f a
100 50
a a
a
b bc
b
de
c
c cd
d ef
0
0 0
2
4 6 Oxidation time days
8
10
e
150
0
2
4 6 Oxidation time days
d fg 8
d e g 10
Figure 2–Phospholipid contents of canola oil emulsions with added soybean lecithin during iron-catalyzed autoxidation (A) or chlorophyll-photosensitized oxidation (B) at 25 °C (x, total phospholipids; , phosphatidylcholine; , phosphatidylethanolamine; •, phosphatidylinositol). Different letters in lines represent significant difference among values within the same kind of phospholipid by Duncan’s multiple range test at α = 0.05. Vol. 79, Nr. 11, 2014 r Journal of Food Science C2207
C: Food Chemistry
Phospholipid class
Oxidation
Lecithin on emulsion oxidation . . .
C: Food Chemistry
and during heating of canola oil (Kim and Choe 2012). Among References sesame seed PLs, PI was shown to be the most stable during mi- An S, Choe E. 2011. Effects of unrefined vegetable oil addition on the flavor acceptability and oil oxidation of tuna oil enriched emulsion under singlet oxygen. Food Sci Biotechnol crowave heating (Yoshida and others 1995). 20:743–50. Under chlorophyll-photosensitized oxidation, a rate constant An S, Lee E, Choe E. 2011. Effects of solubility characteristics of sensitiser and pH on the photooxidation of oil in tuna oil-added acidic O/W emulsions. Food Chem 128:358–63. for the degradation of total soybean PLs was 0.056/d, which was Bosanac I, Michikawa T, Mikoshiba K, Ikura M. 2004. Structural insights into the regulatory not significantly different from that under iron-catalyzed autoxidamechanism of IP3 receptor. Biochim Biophys Acta 1742:89–102. A, Klosinski R, Pawlak A, Wrona-Krol M, Thompson D, Sarna T. 2011. Interaction tion. PI showed a significantly (P < 0.05) higher degradation rate Broniec of plasmalogens and their diacyl analogs with singlet oxygen in selected model systems. Free (0.078/d) than PC (0.053/d) or PE (0.046/d). This result is conRadic Biol Med 50:892–8. flicting with the result in iron-catalyzed oxidation, and could be Carvajal A, Rustad T, Mozuraityte R, Storro I. 2009. Kinetic studies of lipid oxidation induced by hemoglobin measured by consumption of dissolved oxygen in a liposome model system. J attributed to a mechanistic difference between iron-catalyzed oxAgric Food Chem 57:7826–33. idation and chlorophyll-photosensitized oxidation. Radicals such Chen B. 2012. Minor components and their roles on lipid oxidation in bulk oil that contains association colloids [Ph.D. dissertation]. Amherst, Mass.: University. of Massachusetts as alkyl radical and peroxy radical are important factors in ironAmherst. Available from: http://scholarworks.umass.edu/open_access_dissertations/540. Accessed 2014 May 27. catalyzed oxidation, and radical scavenging by hydrogen donation Choe E, Min DB. 2006. Mechanisms and factors for edible oil oxidation. Comp Rev Food Sci is a main action mechanism of PLs as antioxidants. Hydrogen doFood Saf 5:169–86. nation to alkyl radicals of lipids by PLs results in their own degra- Choe J, Choe E. 2013. Effects of chlorophyll addition and light on the oxidative stability and antioxidant changes of perilla oil emulsion. Korean J Food Cookery Sci 29:53–62. dation, and thus PE with a higher hydrogen-donating capability Choi H, Choe E. 2009. Comparison of autoxidative stability and minor compounds in oils extracted from bran and germ of Keumkang wheat and dark northern spring wheat. Korean J than PI could be degraded faster. On the contrary, singlet oxygen Food Sci Technol 41:628–35. production is important in chlorophyll-photosensitized oxidation, Fomuso LB, Corredig M, Akoh CC. 2002. Effect of emulsifier on oxidation properties of fish which may be affected by charged molecules. Charged groups oil-based structured lipid emulsions. J Agric Food Chem 50:2957–61. Garti N. 2002. Food emulsifiers: structure-reactivity relationships, design, and applications. In: in PLs contribute to fast adsorption of PLs to the oil/water inMarangoni AG, Narine S, editors. Physical properties of lipids. New York, N.Y.: Marcel terface, formation of swollen lamellar structures, and electrostatic Dekker, Inc. p 265–386. FJ, Nogales F, Zamora R. 2005. Changes produced in the antioxidative activity of repulsion (Rydhag and Wilton 1981), resulting in a less resistive Hidalgo phospholipids as a consequence of their oxidation. J Agric Food Chem 53:659–62. interfacial film in emulsions. This, in turn, can cause high oxy- Husain SR, Terao J, Matsushita S. 1986. Effect of browning reaction products of phospholipids on autoxidation of methyl linoleate. J Am Oil Chem Soc 63:1457–60. gen diffusion into the oil. Under neutral pH conditions, PC and Jacobson K, Papahadjopoulos D. 1975. Phase transition and phase separations in phospholipid PE act as ampholytes, which contain both a negatively charged membranes induced by changes in temperature, pH and concentration of bivalent cations. Biochem 14:152–61. phosphate group and positively charged choline or ethanolamine, Judde A, Villeneuve P, Rossignol-Castera A, Le Guillou A. 2003. Antioxidant effect of soy whereas PI contains only a negatively charged phosphate group lecithins on vegetable oil stability and their synergism with tocopherols. J Am Oil Chem Soc 80:1209–15. and uncharged inositol (Bosanac and others 2004; Wang and Wang Kim N, Choe E. 2012. Effects of monoacylglycerols on the oil oxidation of acidic water/perilla 2008). Therefore, oxygen diffusion into the oil in emulsion is inoil emulsion under light in the presence of chlorophyll. Food Sci Biotechnol 21:183–9. creased by PI compared to PC or PE, causing higher production Koo B, Kim J. 2005. Effect of individual phospholipid components treating on storaging and frying stability in soybean oil. Korean J Food Culture 20:451–8. of singlet oxygen. This could have resulted in higher contact of PI Kristinov´ a V, Mozuraityte R, Storrø I, Rustad T. 2009. Antioxidant activity of phenolic acids in lipid oxidation catalyzed by different prooxidants. J Agric Food Chem 57:10377–85. than PC or PE with singlet oxygen, causing its faster degradation. a V, Mozuraityte R, Aaneby J, Storrø I, Rustad T. 2014. Iron-mediated peroxidation The antioxidant activity of PC during oxidation of water-in-oil Kristinov´ in marine emulsions and liposomes studied by dissolved oxygen consumption. Eur J Lipid Sci microemulsion in the presence of chemically produced singlet oxyTechnol 116:207–25. J, Choe E. 2009. Effects of phosphatidylcholine and phosphatidylethanolamine on the gen was attributed to the reaction between PC and singlet oxygen, Leephotooxidation of canola oil. J Food Sci 74:C481–6. chemical quenching (Lee and Choe 2008). The singlet oxygen Lee Y, Choe E. 2008. Singlet oxygen quenching effects of phosphatidylcholine in emulsion sunflower oil. J Food Sci 73:C506–11. quenching rate by PC was reported to be 105 /M/s (Broniec and Leecontaining Y, Choe E. 2011. Interaction of phosphatidylcholine and α-tocopherol on the oxidation others 2011). There was no significant difference in degradation of sunflower oil and content changes of phosphatidylcholine and tocopherol in the emulsion under singlet oxygen. J Food Sci 76:C498–503. rate between PC and PE during chlorophyll-photosensitized oxiMozuraityte R, Rustad T, Storrø I. 2006. Pro-oxidant activity of Fe2+ in oxidation of cod dation, which was a similar result to that of wheat bran oil during phospholipids in liposomes. Eur J Lipid Sci Technol 108:218–26. Mozuraityte R, Rustad T, Storro I. 2008. The role of iron in peroxidation of polyunsaturated photooxidation (Choi and Choe 2009).
Conclusions Soybean lecithin added to canola oil emulsions significantly reduced and decelerated iron-catalyzed oil oxidation by lowering headspace oxygen consumption and hydroperoxide production. However, soybean lecithin had no significant antioxidant effect on chlorophyll-photosensitized oxidation of the emulsions. During iron-catalyzed and chlorophyll-photosensitized oxidation, PLs derived from soybean lecithin which was added to the emulsion were degraded, with little change in the PL composition. Among PLs, PE and PI were degraded the fastest in the iron-catalyzed autoxidation and the chlorophyll-photosensitized oxidation, respectively, which was possibly due to radical scavenging and chemical quenching of singlet oxygen, respectively, as a main action mechanism as antioxidants in oil oxidation of the canola oil emulsions.
Acknowledgments This research was supported by the Basic Science Research Program through the Natl. Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT, and Future Planning (NRF-2012R1A1A3011423). C2208 Journal of Food Science r Vol. 79, Nr. 11, 2014
fatty acids in liposomes. J Agric Food Chem 56:537–43. Osborn HT, Akoh CC. 2003. Effects of natural antioxidants on iron-catalyzed lipid oxidation of structured lipid-based emulsions. J Am Oil Chem Soc 80:847–52. Pikul J, Kummerow FA. 1990. Lipid Oxidation in chicken muscles and skin after roasting and refrigerated storage of main broiler parts. J Food Sci 55:30–7. Rydhag L, Wilton I. 1981. The function of phospholipids of soybean lecithin in emulsions. J Am Oil Chem Soc 58:830–7. Saito H, Ishihara K. 1997. Antioxidant activity and active sites of phospholipids as antioxidants. J Am Oil Chem Soc 74:1531–6. Sugino H, Ishikawa M, Nitoda T, Koketsu M, Juneja LR, Kim M, Yamamoto T. 1997. Antioxidative activity of egg yolk phospholipids. J Agric Food Chem 45:551–4. Szuhaj BF. 1989. Lecithins: sources, manufacture and uses. Champaign, Ill.: American Oil Chemists’ Society. p 16–31. Viani P, Cervato G, Florilli A, Rigamonti E, Cestaro B. 1990. Studies on peroxidation of model membranes and synaptosomes: role of phophatidic acid. Chem Phys Lipids 52:49–55. Wang G, Wang T. 2008. Oxidative stability of egg and soy lecithin as affected by transition metal ions and pH in emulsion. J Agric Food Chem 56:11424–31. Xu Q, Nakajima M, Liu Z, Shiina T. 2011. Soybean-based surfactants and their applications. In: Ng T-B, editor. Soybean-applications and technology. Rijeka, Croatia: InTech. Available from: http://www.intechopen.com/books/soybean-applications-and-technology/soybeanbased-surfactants-and-their-applications. Accessed 2014 May 27. Yoon S, Min D. 1987. Roles of phospholipids in flavor stability of soybean oil. Korean J Food Sci Technol 19:23–8. Yoshida H, Shigezaki J, Takagi S, Kajimoto G. 1995. Variations in the composition of various acyl lipids, tocopherols and lignans in sesame seed oils roasted in a microwave oven. J Sci Food Agric 68:407–15. Zambiazi R, Przybylski R. 1998. Effect of endogenous minor components on the oxidative stability of vegetable oils. Lipid Technol 10:58–62.
Abstract: The effect of soybean lecithin addition on the iron-catalyzed or chlorophyll-photosensitized oxidation of emulsions consisting of purified canola oil and water (1:1, w/w) was studied based on headspace oxygen consumption using gas chromatography and hydroperoxide production using the ferric thiocyanate method. Addition levels of iron sulfate, chlorophyll, and soybean lecithin were 5, 4, and 350 mg/kg, respectively. Phospholipids (PLs) during oxidation of the emulsions were monitored by high performance liquid chromatography. Addition of soybean lecithin to the emulsions significantly reduced and decelerated iron-catalyzed oil oxidation by lowering headspace oxygen consumption and hydroperoxide production. However, soybean lecithin had no significant antioxidant effect on chlorophyll-photosensitized oxidation of the emulsions. PLs in soybean lecithin added to the emulsions were degraded during both oxidation processes, although there was little change in PL composition. Among PLs in soybean lecithin, phosphatidylethanolamine and phosphatidylinositol were degraded the fastest in the iron-catalyzed and the chlorophyll-photosensitized oxidation, respectively. The results suggest that addition of soybean lecithin as an emulsifier can also improve the oxidative stability of oil in an emulsion. Keywords: chlorophyll-photosensitized oxidation, degradation of phospholipids, emulsion, iron-catalyzed oxidation, soybean lecithin
The results of this study can be applied to extend the utilization of soybean lecithin for improvement of the oxidative stability of emulsions as well as an emulsifier.
Practical Application:
Introduction The quality of an emulsion is largely determined by the oxidation of oil in the system (Lee and Choe 2011), and the oil oxidation is accelerated by transition metals such as iron as well as pigments such as chlorophylls under light (An and others 2011). An emulsion is stabilized by an emulsifier such as phospholipids (PLs), which have both a polar head and a nonpolar tail in their structures. Until now, there remains a controversy regarding the effect of PL on oil oxidation. Phosphatidylcholine (PC) reduces the autoxidation of bulk oil (Sugino and others 1997; Zambiazi and Przybylski 1998; Koo and Kim 2005) by chelating or binding to prooxidant metal ions (Jacobson and Papahadjopoulos 1975; Viani and others 1990). On the other hand, PL worked as a prooxidant in both autoxidation (Husain and others 1986; Yoon and Min 1987) and photooxidation of bulk oil (Lee and Choe 2009) by increasing oxygen diffusion from the headspace to the oil or protecting chlorophylls, photosensitizers for more reactive singlet oxygen production under light, from photodegradation. 1,2-Dioleoyl-sn-glycerol-3-phosphocholine forming reverse micelles in stripped soybean oil acted as a prooxidant, whereas 1,2-dibutyl-sn-glycerol-3-phosphocholine was not able to form association colloids and did not significantly affect lipid oxidation (Chen 2012). It is generally agreed that the effect of PLs on the oil oxidation is dependent on the type and the concentration of
PLs, a reaction system matrix whether bulk oil or an emulsion, and the presence of iron or other antioxidants (Judde and others 2003; Choe and Min 2006). Most research on the effect of PLs on oil oxidation has examined autoxidation using pure PLs such as pure PC or phosphatidylethanolamine (PE; Jacobson and Papahadjopoulos 1975; Yoon and Min 1987; Sugino and others 1997; Lee and Choe 2009). Lecithin, a common emulsifier used in the food industry, is a mixture of PLs including PC, PE, and phosphatidylinositol (PI), and is naturally found in egg yolk and soybeans. Due to the high cost of lecithin from egg yolk, soybean lecithin is more widely used in the food industry. Since the PL and fatty acid composition of lecithin varies depending on the sources, lecithin extracted from different food materials may have different effects on oil oxidation in an emulsion. Another concern in terms of oil oxidation in emulsion foods with added PLs may be light exposure during transportation and storage, and a photosensitized-oxidation may occur if photosensitizing pigments are present. This research evaluated an oil oxidation and PL degradation in emulsions with added soybean lecithin under iron or light with chlorophyll in order to elucidate the effect and action mechanism of soybean lecithin on iron-catalyzed or chlorophyll-photosensitized oxidation of canola oil emulsions.
Materials and Methods MS 20140955 Submitted 6/3/2014, Accepted 9/5/2014. Authors are with Dept. of Food and Nutrition, Inha Univ., Incheon, South Korea. Direct inquiries to author Choe (E-mail: [email protected]).
R C 2014 Institute of Food Technologists
doi: 10.1111/1750-3841.12683 Further reproduction without permission is prohibited
Materials and reagents Refined, bleached, and deodorized (RBD) canola oil was obtained from CJ Co. (Seoul, Korea), and soybean lecithin was Vol. 79, Nr. 11, 2014 r Journal of Food Science C2203
C: Food Chemistry
Effect of Soybean Lecithin on Iron-Catalyzed or Chlorophyll-Photosensitized Oxidation of Canola Oil Emulsion
Lecithin on emulsion oxidation . . .
C: Food Chemistry
purchased from Sigma-Aldrich Co. (St. Louis, Mo., U.S.A.). n-Hexane, methanol, acetonitrile, and chloroform were high performance liquid chromatography (HPLC) grade and purchased from Mallinckrodt Baker Co. (Phillipsburg, N.J., U.S.A.). Chlorophyll b, 14% BF3 in methanol, silicic acid, alumina, xanthan gum, standard PC, PE, PI, phosphatidylserine (PS), heptadecanoic acid, methyl esters of standard fatty acids (C4–C24 saturated fatty acids [SFAs] as well as oleic, linoleic, and α-linolenic acids), and ammonium thiocyanate were obtained from Sigma-Aldrich Co. Isooctane, celite, and ferrous sulfate were purchased from Junsei Co. (Tokyo, Japan). All other chemicals were analytical grade.
Preparation of emulsions Emulsion samples were prepared with tocopherol-stripped canola oil (TSCO), distilled water, and xanthan gum according to the method of Lee and Choe (2011). Canola oil was chosen due to its relatively high content of linolenic acid (11.48 ± 0.05%), one of the omega-3 fatty acids. TSCO was obtained by passing the RBD canola oil through a glass column (30 cm × 2.5 cm) packed with activated silicic acid, alumina, and celite from the bottom (Lee and Choe 2009). TSCO (49.8 g), water (49.8 g), and xanthan gum (0.35 g) were mixed at room temperature, followed by addition of FeSO4 or chlorophyll b, and soybean lecithin. Concentration of FeSO4, chlorophyll b, and soybean lecithin were 5, 4, and 350 mg/kg, respectively, which were within the range to show significant effect on the oil oxidation in emulsions (Osborn and Akoh 2003; Lee and Choe 2011). The mixture was then homogenized for 6 min using an Ultra-Turrax T25 homogenizer equipped with an S25N-25F dispersing tool (IKA Instruments, Staufen, Germany). Control samples did not contain soybean lecithin, but the emulsion system was maintained for 10 d similar to other samples, possibly due to xanthan gum. The emulsion (3 g) was transferred into 20-mL serum bottles, which were then tightly capped with rubber stoppers and aluminum caps. Samples with added iron were covered with aluminum foil to exclude light. All samples in bottles were oxidized in an LBI-250 incubator (Daihan Labtech Co., Seoul, Korea) under 1700 lux light and maintained at 25 °C for 10 d until taken out for analyses. All samples were prepared in duplicate. Evaluation of oil oxidation in emulsions Oil oxidation in the emulsions was determined based on the headspace oxygen consumption as a loss of reactant (oxygen) in the oil oxidation, and the hydroperoxide contents were also determined as a primary product of oil oxidation. Monitoring of oxygen consumption is a very useful tool for the analysis of oil oxidation in an emulsion since it does not require extracting the oil phase before analysis and focuses on the loss of substrate in the oxidation reactions (Kristinov´a and others 2009). It also enables continuous monitoring of the oxygen concentration in the system and easy changes of environmental conditions such as light as well as addition of prooxidants and antioxidants to the system during the oxidation process. Thus, the oxygen consumption has been employed to provide direct information on the oil oxidation in emulsions (Mozuraityte and others 2006; Mozuraityte and others 2008; Carvajal and others 2009; Kristinov´a and others 2009; An and Choe 2011; Kristinova and others 2014). Headspace oxygen content was determined by gas chromatography (GC; Lee and Choe 2011) using a YL 6100 gas chromatograph (Younglin Instrument Co., Ltd., Anyang, Korea) equipped with an autosampler and a thermal conductivity detector. The injection volume was 0.5 mL. A column was the stainless steel column packed with C2204 Journal of Food Science r Vol. 79, Nr. 11, 2014
80/100 mesh molecular sieve 13× (1.83 m × 0.32 cm; Alltech, Deerfield, Ill., U.S.A.). Helium (99.995%) was a carrier gas at 20 mL/min. Temperatures of the oven, injector, and detector were 35, 100, and 140 °C, respectively. Areas of the oxygen peak in the chromatograms were converted to μmol of oxygen in 1 mL of headspace gas with a conversion factor of 9.35 (Lee and Choe 2011). Hydroperoxide content of the emulsion was determined by the ferric thiocyanate method (An and Choe 2011). After mixing the emulsion with isooctane and 2-propanol mixture (3:1, v/v), the organic layer was separated by centrifugation at 1000 × g for 2 min, followed by addition of methanol and chloroform mixture (2:1, v/v), 3.94 mol/L ammonium thiocyanate, 0.132 mol/L BaCl2 , and 0.144 mol/L FeSO4 solution in the respective order. The absorbance at 510 nm using a HP8453 UV-visible spectrophotometer (Hewlett Packard, Wilmington, Del., U.S.A.) was read after 20 min. Hydroperoxide content of oil was expressed with cumene hydroperoxide (CuOOH), and the range of linearity was between 0 and 10 mmol/kg (r2 = 0.9994).
Compositional analyses of soybean lecithin PL composition of soybean lecithin was analyzed by HPLC (Choe and Choe 2013). Soybean lecithin was dissolved in chloroform and then injected into a YL 9100 HPLC (Younglin Instrument Co., Ltd.) equipped with a UV detector at 205 nm. The injection volume was 20 μL. A cosmosil-packed column (5SL-II; 4.6 mm × 250 mm, 5 μm size; Nacalai Tesque, Kyoto, Japan) was used with an eluting solvent of acetonitrile, methanol, and phosphoric acid (130:5:1.5, v/v/v) at 1.0 mL/min in an isocratic manner. Each PL in soybean lecithin was identified by comparing the retention time with those of standard PC, PE, PI, and PS, and quantified with their respective calibration curves with a linearity range between 0 and 400 mg/L (r2 > 0.9950). Fatty acid composition of soybean lecithin was analyzed by GC according to the method of Lee and Choe (2011) after the lecithin was saponified with methanolic NaOH (2 mol/L) and esterified with 14% BF3 in methanol. The instrument was a YL 6100 gas chromatograph (Younglin Instrument Co., Ltd.) equipped with an autosampler, an HP-Innowax column (30 m × 0.53 mm, 1.0-μm thick; Agilent, B¨oblingen, Germany), and a flame ionization detector. The injection volume was 1 μL. The temperatures of the oven, injector, and detector were 200, 270, and 280 °C, respectively. The helium flow rate was 10 mL/min, and the split ratio was 10:1. Heptadecanoic acid was used as an internal standard, and each fatty acid in the chromatogram was identified by comparing retention times with those of standard fatty acid methyl esters. Concentration of each fatty acid was represented as relative % by dividing each peak area by total peak areas. Determination of PL contents in emulsions during oxidation Content of each PL in the emulsion during 10-d oxidation was determined by HPLC after phase separation. The emulsion was frozen, thawed, and centrifuged, and both interfacial and water phases were dissolved in n-hexane. The upper layer obtained by centrifugation (2000 × g, 10 min) was applied to a solid phase microextraction Sep-pak silica plus long cartridge (55–105 μm; Waters Corp. Milford, Mass., U.S.A.), and eluted with chloroform and methanol in a respective order (Lee and Choe 2011). PLs eluted with methanol were analyzed by HPLC using a YL 9100 HPLC (Younglin Instrument Co., Ltd.) under the same analytical conditions described previously.
Lecithin on emulsion oxidation . . .
PL content (mg/100 mg lecithin)
Fatty acid compositionb
Phosphatidylcholine
Phosphatidylethanolamine
Phosphatidylinositol
Phosphatidylserine
Total
31.9 ± 0.2 (40.7)a
26.4 ± 0.1 (33.6)
18.0 ± 0.2 (23.0)
2.1 ± 0.0 (2.7)
78.4 ± 0.1 (100)
Palmitic acid
Stearic acid
Oleic acid
Linoleic acid
Linolenic acid
USFA/SFAc
21.2 ± 0.2
4.2 ± 0.1
16.2 ± 0.1
53.4 ± 0.1
5.1 ± 0.1
3.0 ± 0.0
a
Relativity in % based on total PL content. b Relativity in % based on total fatty acids. c Content ratio of unsaturated fatty acids to saturated fatty acids.
Data analysis All samples were prepared in duplicate, and each sample was measured twice, resulting in 4 measurements for each treatment. Data were analyzed using one-way analysis of variance and Duncan’s multiple range test using SAS (version 9.2, SAS Inst. Inc., Cary, N.C., U.S.A.) at α = 0.05.
Results and Discussion
12
7 a b
6 c c
5
a
Hydroperoxide contents mmol CuOOH/ kg
Headspace oxygen consumption μmol O2 / mL
Composition of soybean lecithin PL content of soybean lecithin was 78.4%, and PL consisted of mainly PC (40.7%), PE (33.6%), and PI (23.0%) with a small amount of PS (2.7%), as shown in Table 1. These values were within the ranges reported by Garti (2002). Linoleic acid (53.4%) was the most abundant fatty acid in soybean lecithin along with palmitic (21.2%) and oleic (16.2%) acids. Linolenic (5.1%) and stearic (4.2%) acids were present at low concentration, resulting in the content ratio of unsaturated fatty acids (USFA) to SFAs of 3.0. High unsaturation of soybean lecithin was reported elsewhere (Szuhaj 1989), and high unsaturation of PL was suggested for its prooxidant effect in the oxidation of chicken fat (Pikul and Kummerow 1990).
Effect of soybean lecithin addition on oil oxidation in canola oil emulsions Headspace oxygen consumption and hydroperoxide contents of the canola oil emulsions increased during both iron-catalyzed autoxidation and chlorophyll-photosensitized oxidation at 25 °C for 10 d, as shown in Figure 1. The control emulsion without added soybean lecithin showed oxygen consumption levels of 1.14, 3.62, and 5.41 μmol O2 /mL as well as hydroperoxide contents of 2.55, 4.66, and 5.71 mmol CuOOH/kg after 2, 6, and 10 d of iron-catalyzed oxidation, respectively. Oxygen consumption and hydroperoxide contents of the emulsion with added soybean lecithin after 10 d of iron-catalyzed oxidation were significantly reduced to 3.32 μmol O2 /mL and 3.16 mmol CuOOH/kg, respectively, which were significantly lower than those of the control emulsion (P < 0.05). These results indicate that soybean lecithin significantly reduced oil oxidation catalyzed by iron in the canola oil emulsions. Added soybean lecithin has been reported to show antioxidant activity during heating of vegetable oil at 110 °C (Judde and others 2003), whereas Fomuso and others (2002) reported increased autoxidation of fish oil-based structured lipid emulsion by addition of lecithin.
c d
4 ef f
3
e ef g
g 2
h i
1 j 0
h i
j
kj 0
2
4 6 Oxidation time days
8
10
a
10
b 8
bc
c d
6
e
e
f 4 gh gh h
2
i
i
0 0
2
g
g gh h
h i 4 6 Oxidation time days
8
10
Figure 1–Effect of soybean lecithin (350 mg/kg) on the headspace oxygen consumption and hydroperoxide contents of canola oil emulsions during iron-catalyzed autoxidation (solid line) or chlorophyll-photosensitized oxidation (dotted line) at 25 °C (x, no soybean lecithin addition (control); , with soybean lecithin addition). Different letters in lines represent significant difference among values regardless of oxidation time and treatment in each oxidation parameter by Duncan’s multiple range test at α = 0.05. Vol. 79, Nr. 11, 2014 r Journal of Food Science C2205
C: Food Chemistry
Table 1–Composition of soybean lecithin.
Lecithin on emulsion oxidation . . . Table 2–Effect of soybean lecithin addition on regression analysis between oxidation time and headspace oxygen consumption or hydroperoxide contents of canola oil emulsions during iron-catalyzed autoxidation or chlorophyll-photosensitized oxidation at 25 °C for 10 d. Oxidation
Regression parametersa
Soybean lecithin addition
Headspace oxygen consumption (μmol O2 /mL)
C: Food Chemistry
Iron-catalyzed Chlorophyll-photosensitized
No (control) Yes No (control) Yes
Hydroperoxide contents (mmol CuOOH/kg)
a
b
r2
a
b
r2
0.556 ab Ac 0.340 bb 0.710 ab Ac 0.685 ab
0.098 −0.198 0.190 0.498
0.994 0.989 0.947 0.955
0.582 ab Ac 0.319 bb 0.976 ab Ac 0.908 ab
0.535 −0.335 0.245 0.447
0.940 0.918 0.864 0.901
Headspace oxygen consumption (μmol O2 /mL) or hydroperoxide contents (mmol CuOOH/kg) = a × oxidation time (d) + b, r2 = determination coefficient Different lowercase letters represent a significant difference in the slope between control and sample with added soybean lecithin under each oxidation by dummy variable regression analysis (P < 0.05). c Same uppercase letters represent no significant difference in the slope between control samples under iron-catalyzed autoxidation and chlorophyll-photosensitized oxidation by dummy variable regression analysis (P < 0.05). a
b
Oxygen consumption and hydroperoxide production of the control emulsion under chlorophyll-photosensitized oxidation were significantly higher than those under iron-catalyzed autoxidation (P < 0.05); headspace oxygen consumption of the control emulsion without soybean lecithin addition after 6 and 10 d of oxidation was 5.46 and 6.53 μmol O2 /mL, while hydroperoxide content was 6.63 and 8.16 mmol CuOOH/kg, respectively. This result clearly confirms higher oil oxidation in the emulsions under light in the presence of chlorophyll than in the dark in the iron presence. Chlorophyll acts as a photosensitizer for the production of singlet oxygen, whose reactivity with lipids is approximately 1400 times higher than that of triplet oxygen (Choe and Min 2006). Unlike iron-catalyzed oxidation, addition of soybean lecithin did not significantly reduce oxygen consumption or hydroperoxide production in the emulsions under chlorophyll-photosensitized oxidation. Oxygen consumption and hydroperoxide content of the emulsion with added soybean lecithin after 2, 6, and 10 d of oxidation were 1.92 μmol O2 /mL and 2.47 mmol CuOOH/kg, 5.19 μmol O2 /mL and 7.57 mmol CuOOH/kg, and 6.52 μmol O2 /mL and 8.22 mmol CuOOH/kg, respectively, which were not significantly different from those of the control emulsion (P > 0.05). These results indicate that soybean lecithin did not affect chlorophyll-photosensitized oxidation of the canola oil emulsions. Until now, the effect of PLs on chlorophyll-photosensitized oxidation in emulsions has not been reported. Headspace oxygen consumption and hydroperoxide production in the canola oil emulsions were highly correlated with the oxidation time (r2 > 0.86), as shown in Table 2. Rates of oil oxidation were estimated based on the slope of the regression lines between time and oxygen consumption or hydroperoxide content, as previously described (Kim and Choe 2012). The rate of oil oxidation in the control emulsion was significantly higher (P < 0.05) under chlorophyll-photosensitized oxidation (0.710 μmol O2 /mL/d and 0.976 mmol CuOOH/kg/d) than under iron-catalyzed oxidation (0.556 μmol O2 /mL/d and 0.582 mmol CuOOH/kg/d). Further, the rate of oil oxidation was significantly lower (P < 0.05) in the emulsion with added soybean lecithin (0.340 μmol O2 /mL/d and 0.319 mmol CuOOH/kg/d) than in the control emulsion only under iron-catalyzed autoxidation. In the presence of chlorophyll under light, the rate of oil oxidation in the emulsion with added soybean lecithin was 0.685 μmol O2 /mL/d based on oxygen consumption and 0.908 mmol CuOOH/kg/d based on the hydroperoxide production, and these were not significantly different from the rate of oil oxidation in the
C2206 Journal of Food Science r Vol. 79, Nr. 11, 2014
control emulsion (P > 0.05). PLs sequester metals and/or donate hydrogen atoms from their nitrogen-containing functional groups such as choline or ethanolamine to radicals (Saito and Ishihara 1997), resulting in decreased oil oxidation (Choe and Min 2006). On the other hand, PLs can elevate autoxidation and chlorophyllphotosensitized oxidation of purified oil which does not have any metals (Yoon and Min 1987; Lee and Choe 2009). Since PLs are located mainly at interfaces with their nonpolar fatty acid tails facing the oil phase and a polar head facing the water phase (Xu and others 2011), they are able to decrease the surface tension of oil and increase oxygen diffusion from the headspace, resulting in acceleration of oil oxidation (Choe and Min 2006). Therefore, the reduction of iron-catalyzed autoxidation of oil by added soybean lecithin could have been largely attributed to iron chelation by PLs. Iron increases oil oxidation by accelerating decomposition of hydroperoxides into radicals (Choe and Min 2006). In addition, donation of hydrogen to radicals from functional groups such as choline or ethanolamine could have contributed to the reduction of oil oxidation. Although PLs can also donate hydrogens to radicals under chlorophyll-photosensitized oxidation, increased oxygen diffusion by PLs could have resulted in a higher singlet oxygen production, which was more predominant than radical scavenging by hydrogen donation in chlorophyll-photosensitized oxidation. Thus, addition of soybean lecithin had no significant effect on the chlorophyll-photosensitized oxidation of the canola oil emulsions.
PL contents in emulsions with added soybean lecithin during oxidation PL content of the emulsion with added soybean lecithin was measured to be 252.76 mg/kg before oxidation, which was lower than the expected level (274.4 mg/kg) since soybean lecithin with 78.40% of purity was added at 350 mg/kg. This suggests that some of PL in soybean lecithin were lost during PL analysis, with 92.1% of recovery. Content of PC (100.88 mg/kg), PE (90.26 mg/kg), and PI (61.62 mg/kg) corresponded to 39.9, 35.7, and 24.4%, respectively, of total PL content. There was no PS detected in the emulsions, possibly due to a very low amount compared to other PLs. During iron-catalyzed or chlorophyllphotosensitized oxidation at 25°C, PL contents in the canola oil emulsions significantly decreased (Figure 2), indicating PL degradation. However, PL composition changed little, with a slight decrease in the relative content of PE and an increase in PI. After 10 d of oxidation catalyzed by iron, contents of PC, PE, and PI
Lecithin on emulsion oxidation . . . Table 3–Degradation of phospholipids in canola oil emulsions with added soybean lecithin during iron-catalyzed autoxidation or chlorophyll-photosensitized oxidation at 25 o C for 10 d.
Iron-catalyzed
Regression parametersa
Phosphatidylcholine Phosphatidylethanolamine Phosphatidylinositol Total Phosphatidylcholine Phosphatidylethanolamine Phosphatidylinositol Total
Chlorophyll-photosensitized
k
t1/2
r2
0.065bb 0.094a 0.044bc 0.068Ac 0.053bb 0.046bc 0.078a 0.056Ac
10.664 7.374 15.753 10.193 13.078 15.068 8.887 12.378
0.9704 0.8940 0.8257 0.9656 0.9462 0.9061 0.9953 0.9723
Estimated by regression assuming 1st-order kinetics, ln([A]/[A]0 = −k × time (d), where [A] and [A]0 are contents of phospholipids (mg/kg) at times t and 0, respectively. r2 , determination coefficient; t1/2 , half-life. b Different letters (a, b, c) represent significant difference in degradation rate among phospholipid classes under each oxidation by dummy variable regression analysis (P < 0.05). c Same letters (A, A) represent no significant difference in degradation rates of total phospholipids between iron-catalyzed autoxidation and chlorophyll-photosensitized oxidation by dummy variable regression analysis (P < 0.05). a
were 51.51, 38.04, and 37.33 mg/kg, respectively, corresponding to 40.6%, 30.0%, and 29.4% of the initial content, giving the total PL content of 126.88 mg/kg (50.2% of the initial content). Degradation of PLs is related to their antioxidant activity by hydrogen donation to lipids (Yoshida and others 1995), and PC and PE were reported to be degraded into carbonyl groups and pyrrole compounds during autoxidation of soybean oil (Hidalgo and others 2005). PLs in the canola oil emulsions with added soybean lecithin were also degraded during chlorophyll-photosensitized oxidation at 25 °C (Figure 2). After an oxidation catalyzed by chlorophyll under light, contents of PC, PE, and PI significantly decreased to 41.3%, 35.5%, and 23.2% (after 2 d), 39.3%, 39.3%, and 21.4% (after 6 d), and 41.5%, 38.2%, and 20.3% (after 10 d) of the initial content, respectively. Total PL contents decreased to 234.07, 180.04, and 142.86 mg/kg after 2, 6, and 10 d of oxidation,
respectively, however, PL composition of the emulsions changed little, similar to the case of iron-catalyzed oxidation. Degradation of PLs in the canola oil emulsion with added soybean lecithin during iron-catalyzed or chlorophyll-photosensitized oxidation followed a 1st-order kinetics (r2 > 0.82). Rate constants (k value) for the 1st-order degradation of PLs were estimated by regression of the natural log (ln) of the PL content ratio ([A]t /[A]0 , where [A]0 and [A]t are PL contents at oxidation times 0 and t, respectively) against time, as shown in Table 3. A rate constant for the degradation of total soybean PLs during iron-catalyzed oxidation of the emulsion was 0.068/d, and the rate constants for PC, PE, and PI degradation were 0.065, 0.094, and 0.044/d, respectively. Among soybean PLs, degradation of PE was significantly faster (P < 0.05) than that of PC or PI during iron-catalyzed oxidation. Lower stability of PE compared to PC was reported during autoxidation of wheat bran oil (Choi and Choe 2009)
B Chlorophyllphotosensitized oxidation
A Iron-catalyzed autoxidation
350
350 300
a
250
b
200
c
cd f
150
g
a 100 50
Phospholipid contents mg/ kg
Phospholipid contents mg/ kg
300
a ab
b a ab
c
c
c
c cd cd
d cd
f
d g ef
a
250
b
200
c
d
f a
100 50
a a
a
b bc
b
de
c
c cd
d ef
0
0 0
2
4 6 Oxidation time days
8
10
e
150
0
2
4 6 Oxidation time days
d fg 8
d e g 10
Figure 2–Phospholipid contents of canola oil emulsions with added soybean lecithin during iron-catalyzed autoxidation (A) or chlorophyll-photosensitized oxidation (B) at 25 °C (x, total phospholipids; , phosphatidylcholine; , phosphatidylethanolamine; •, phosphatidylinositol). Different letters in lines represent significant difference among values within the same kind of phospholipid by Duncan’s multiple range test at α = 0.05. Vol. 79, Nr. 11, 2014 r Journal of Food Science C2207
C: Food Chemistry
Phospholipid class
Oxidation
Lecithin on emulsion oxidation . . .
C: Food Chemistry
and during heating of canola oil (Kim and Choe 2012). Among References sesame seed PLs, PI was shown to be the most stable during mi- An S, Choe E. 2011. Effects of unrefined vegetable oil addition on the flavor acceptability and oil oxidation of tuna oil enriched emulsion under singlet oxygen. Food Sci Biotechnol crowave heating (Yoshida and others 1995). 20:743–50. Under chlorophyll-photosensitized oxidation, a rate constant An S, Lee E, Choe E. 2011. Effects of solubility characteristics of sensitiser and pH on the photooxidation of oil in tuna oil-added acidic O/W emulsions. Food Chem 128:358–63. for the degradation of total soybean PLs was 0.056/d, which was Bosanac I, Michikawa T, Mikoshiba K, Ikura M. 2004. Structural insights into the regulatory not significantly different from that under iron-catalyzed autoxidamechanism of IP3 receptor. Biochim Biophys Acta 1742:89–102. A, Klosinski R, Pawlak A, Wrona-Krol M, Thompson D, Sarna T. 2011. Interaction tion. PI showed a significantly (P < 0.05) higher degradation rate Broniec of plasmalogens and their diacyl analogs with singlet oxygen in selected model systems. Free (0.078/d) than PC (0.053/d) or PE (0.046/d). This result is conRadic Biol Med 50:892–8. flicting with the result in iron-catalyzed oxidation, and could be Carvajal A, Rustad T, Mozuraityte R, Storro I. 2009. Kinetic studies of lipid oxidation induced by hemoglobin measured by consumption of dissolved oxygen in a liposome model system. J attributed to a mechanistic difference between iron-catalyzed oxAgric Food Chem 57:7826–33. idation and chlorophyll-photosensitized oxidation. Radicals such Chen B. 2012. Minor components and their roles on lipid oxidation in bulk oil that contains association colloids [Ph.D. dissertation]. Amherst, Mass.: University. of Massachusetts as alkyl radical and peroxy radical are important factors in ironAmherst. Available from: http://scholarworks.umass.edu/open_access_dissertations/540. Accessed 2014 May 27. catalyzed oxidation, and radical scavenging by hydrogen donation Choe E, Min DB. 2006. Mechanisms and factors for edible oil oxidation. Comp Rev Food Sci is a main action mechanism of PLs as antioxidants. Hydrogen doFood Saf 5:169–86. nation to alkyl radicals of lipids by PLs results in their own degra- Choe J, Choe E. 2013. Effects of chlorophyll addition and light on the oxidative stability and antioxidant changes of perilla oil emulsion. Korean J Food Cookery Sci 29:53–62. dation, and thus PE with a higher hydrogen-donating capability Choi H, Choe E. 2009. Comparison of autoxidative stability and minor compounds in oils extracted from bran and germ of Keumkang wheat and dark northern spring wheat. Korean J than PI could be degraded faster. On the contrary, singlet oxygen Food Sci Technol 41:628–35. production is important in chlorophyll-photosensitized oxidation, Fomuso LB, Corredig M, Akoh CC. 2002. Effect of emulsifier on oxidation properties of fish which may be affected by charged molecules. Charged groups oil-based structured lipid emulsions. J Agric Food Chem 50:2957–61. Garti N. 2002. Food emulsifiers: structure-reactivity relationships, design, and applications. In: in PLs contribute to fast adsorption of PLs to the oil/water inMarangoni AG, Narine S, editors. Physical properties of lipids. New York, N.Y.: Marcel terface, formation of swollen lamellar structures, and electrostatic Dekker, Inc. p 265–386. FJ, Nogales F, Zamora R. 2005. Changes produced in the antioxidative activity of repulsion (Rydhag and Wilton 1981), resulting in a less resistive Hidalgo phospholipids as a consequence of their oxidation. J Agric Food Chem 53:659–62. interfacial film in emulsions. This, in turn, can cause high oxy- Husain SR, Terao J, Matsushita S. 1986. Effect of browning reaction products of phospholipids on autoxidation of methyl linoleate. J Am Oil Chem Soc 63:1457–60. gen diffusion into the oil. Under neutral pH conditions, PC and Jacobson K, Papahadjopoulos D. 1975. Phase transition and phase separations in phospholipid PE act as ampholytes, which contain both a negatively charged membranes induced by changes in temperature, pH and concentration of bivalent cations. Biochem 14:152–61. phosphate group and positively charged choline or ethanolamine, Judde A, Villeneuve P, Rossignol-Castera A, Le Guillou A. 2003. Antioxidant effect of soy whereas PI contains only a negatively charged phosphate group lecithins on vegetable oil stability and their synergism with tocopherols. J Am Oil Chem Soc 80:1209–15. and uncharged inositol (Bosanac and others 2004; Wang and Wang Kim N, Choe E. 2012. Effects of monoacylglycerols on the oil oxidation of acidic water/perilla 2008). Therefore, oxygen diffusion into the oil in emulsion is inoil emulsion under light in the presence of chlorophyll. Food Sci Biotechnol 21:183–9. creased by PI compared to PC or PE, causing higher production Koo B, Kim J. 2005. Effect of individual phospholipid components treating on storaging and frying stability in soybean oil. Korean J Food Culture 20:451–8. of singlet oxygen. This could have resulted in higher contact of PI Kristinov´ a V, Mozuraityte R, Storrø I, Rustad T. 2009. Antioxidant activity of phenolic acids in lipid oxidation catalyzed by different prooxidants. J Agric Food Chem 57:10377–85. than PC or PE with singlet oxygen, causing its faster degradation. a V, Mozuraityte R, Aaneby J, Storrø I, Rustad T. 2014. Iron-mediated peroxidation The antioxidant activity of PC during oxidation of water-in-oil Kristinov´ in marine emulsions and liposomes studied by dissolved oxygen consumption. Eur J Lipid Sci microemulsion in the presence of chemically produced singlet oxyTechnol 116:207–25. J, Choe E. 2009. Effects of phosphatidylcholine and phosphatidylethanolamine on the gen was attributed to the reaction between PC and singlet oxygen, Leephotooxidation of canola oil. J Food Sci 74:C481–6. chemical quenching (Lee and Choe 2008). The singlet oxygen Lee Y, Choe E. 2008. Singlet oxygen quenching effects of phosphatidylcholine in emulsion sunflower oil. J Food Sci 73:C506–11. quenching rate by PC was reported to be 105 /M/s (Broniec and Leecontaining Y, Choe E. 2011. Interaction of phosphatidylcholine and α-tocopherol on the oxidation others 2011). There was no significant difference in degradation of sunflower oil and content changes of phosphatidylcholine and tocopherol in the emulsion under singlet oxygen. J Food Sci 76:C498–503. rate between PC and PE during chlorophyll-photosensitized oxiMozuraityte R, Rustad T, Storrø I. 2006. Pro-oxidant activity of Fe2+ in oxidation of cod dation, which was a similar result to that of wheat bran oil during phospholipids in liposomes. Eur J Lipid Sci Technol 108:218–26. Mozuraityte R, Rustad T, Storro I. 2008. The role of iron in peroxidation of polyunsaturated photooxidation (Choi and Choe 2009).
Conclusions Soybean lecithin added to canola oil emulsions significantly reduced and decelerated iron-catalyzed oil oxidation by lowering headspace oxygen consumption and hydroperoxide production. However, soybean lecithin had no significant antioxidant effect on chlorophyll-photosensitized oxidation of the emulsions. During iron-catalyzed and chlorophyll-photosensitized oxidation, PLs derived from soybean lecithin which was added to the emulsion were degraded, with little change in the PL composition. Among PLs, PE and PI were degraded the fastest in the iron-catalyzed autoxidation and the chlorophyll-photosensitized oxidation, respectively, which was possibly due to radical scavenging and chemical quenching of singlet oxygen, respectively, as a main action mechanism as antioxidants in oil oxidation of the canola oil emulsions.
Acknowledgments This research was supported by the Basic Science Research Program through the Natl. Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT, and Future Planning (NRF-2012R1A1A3011423). C2208 Journal of Food Science r Vol. 79, Nr. 11, 2014
fatty acids in liposomes. J Agric Food Chem 56:537–43. Osborn HT, Akoh CC. 2003. Effects of natural antioxidants on iron-catalyzed lipid oxidation of structured lipid-based emulsions. J Am Oil Chem Soc 80:847–52. Pikul J, Kummerow FA. 1990. Lipid Oxidation in chicken muscles and skin after roasting and refrigerated storage of main broiler parts. J Food Sci 55:30–7. Rydhag L, Wilton I. 1981. The function of phospholipids of soybean lecithin in emulsions. J Am Oil Chem Soc 58:830–7. Saito H, Ishihara K. 1997. Antioxidant activity and active sites of phospholipids as antioxidants. J Am Oil Chem Soc 74:1531–6. Sugino H, Ishikawa M, Nitoda T, Koketsu M, Juneja LR, Kim M, Yamamoto T. 1997. Antioxidative activity of egg yolk phospholipids. J Agric Food Chem 45:551–4. Szuhaj BF. 1989. Lecithins: sources, manufacture and uses. Champaign, Ill.: American Oil Chemists’ Society. p 16–31. Viani P, Cervato G, Florilli A, Rigamonti E, Cestaro B. 1990. Studies on peroxidation of model membranes and synaptosomes: role of phophatidic acid. Chem Phys Lipids 52:49–55. Wang G, Wang T. 2008. Oxidative stability of egg and soy lecithin as affected by transition metal ions and pH in emulsion. J Agric Food Chem 56:11424–31. Xu Q, Nakajima M, Liu Z, Shiina T. 2011. Soybean-based surfactants and their applications. In: Ng T-B, editor. Soybean-applications and technology. Rijeka, Croatia: InTech. Available from: http://www.intechopen.com/books/soybean-applications-and-technology/soybeanbased-surfactants-and-their-applications. Accessed 2014 May 27. Yoon S, Min D. 1987. Roles of phospholipids in flavor stability of soybean oil. Korean J Food Sci Technol 19:23–8. Yoshida H, Shigezaki J, Takagi S, Kajimoto G. 1995. Variations in the composition of various acyl lipids, tocopherols and lignans in sesame seed oils roasted in a microwave oven. J Sci Food Agric 68:407–15. Zambiazi R, Przybylski R. 1998. Effect of endogenous minor components on the oxidative stability of vegetable oils. Lipid Technol 10:58–62.
